15 May 2023 in Medical Image Analysis
1 February 2020 in Cancer Management and Research
Objective: Ascites, an accumulation of peritoneal fluid, is associated with poor prognosis of certain cancers. The potential mechanism that ascites worsens prognosis has not been well understood. Lipids have been reported to correlate with the prognosis of patients with epithelial ovarian cancer (EOC). Therefore, we aimed here to investigate whether lipids mediate the effect of ascites on the recurrence of EOC. Methods: We collected the demographic and pathological data of 437 previously untreated patients with EOC to investigate the influence of ascites on recurrence. To identify the mechanism that mediates the potential influence of ascites on recurrence, we used ultra-performance liquid chromatography coupled with mass spectrometry (UPLC-MS) to determine the plasma lipid profiles of 53 patients with EOC. We used mediation analysis to evaluate if lipids mediated the effects of ascites on the recurrence of EOC. Results: Patients with ascites had a poorer prognosis, which was associated with higher levels of carbohydrate antigen-CA125 (CA125) and FIGO stage. We identified six different lipid metabolites that were associated with ascites and recurrence. Mediation analysis revealed that the lipids LysoPC(P-15:0), PC(P-34:4), and PC(38:6) may mediate the effects of ascites on recurrence. Conclusion: Our findings suggest that LysoPC(P-15:0), PC(P-34:4), and PC(38:6) mediate the effect of ascites on the prognosis of patients with EOC. We believe therefore that it is reasonable to consider metabolic interventions targeting the metabolism of LysoPC(P-15:0), PC(P-34:4), and PC(38:6) as a palliative treatment for patients with EOC with ascites. Further studies of more patients will be required to validate our findings.
23 September 2024 in Briefings in Bioinformatics
The integration of data from multiple modalities generated by single-cell omics technologies is crucial for accurately identifying cell states. One challenge in comprehending multi-omics data resides in mosaic integration, in which different data modalities are profiled in different subsets of cells, as it requires simultaneous batch effect removal and modality alignment. Here, we develop Multi-omics Mosaic Auto-scaling Attention Variational Inference (mmAAVI), a scalable deep generative model for single-cell mosaic integration. Leveraging auto-scaling self-attention mechanisms, mmAAVI can map arbitrary combinations of omics to the common embedding space. If existing well-annotated cell states, the model can perform semisupervised learning to utilize existing these annotations. We validated the performance of mmAAVI and five other commonly used methods on four benchmark datasets, which vary in cell numbers, omics types, and missing patterns. mmAAVI consistently demonstrated its superiority. We also validated mmAAVI's ability for cell state knowledge transfer, achieving balanced accuracies of 0.82 and 0.97 with less 1% labeled cells between batches with completely different omics. The full package is available at https://github.com/luyiyun/mmAAVI.